Organizer(s):

Raf Frongillo, University of Colorado

David Pennock, DIMACS

Jens Witkowski, Frankfurt School of Finance & Management

Description

Following the successful iterations at EC 2017 and DIMACS 2021, we seek submissions to the DIMACS 2024 Workshop on Forecasting. The workshop will be held on October 14, 2024, at Rutgers University New Brunswick, NJ, co-located with the 8th International Conference on Algorithmic Decision Theory (ADT 2024). We welcome submissions describing recent research on crowd-sourced, data-driven, or hybrid approaches to forecasting.

In this workshop, we will bring together computer scientists, economists, statisticians, and decision scientists, some who develop theories of forecasting and others who study it empirically. We invite academics together with practitioners who build forecasting platforms, operate forecasting competitions, and publish predictions. Our primary focus is on what happens after predictive models have been trained or formed; that said, still in scope are data-driven and machine-learning-based techniques that aggregate forecasts and other information to harness the so-called wisdom of the crowd.

Topics of interest for the workshop include but are not limited to:

1. Incentives in forecasting: methods for eliciting truthful and accurate forecasts or information. Incentive mechanisms for information gathering; monetary rewards, competition mechanisms, and implicit reward mechanisms, in both online and batch settings.
2. Forecast evaluation: methods to identify elite forecasters for accuracy-weighted averages or smaller, more selective crowds; evaluating complex probabilities including conditionals, continuous random variables, and exponentially large joint distributions.
3. Forecast aggregation: combining multiple forecasts, including crowdsourced human judgments and data-driven predictions.
4. Prediction markets and related mechanisms: design of prediction market architectures, automated market makers, design of options and derivatives for eliciting information, extracting information from existing financial and gambling markets; wagering mechanisms and other group forecasting mechanisms.
5. Behavioral aspects of forecasting: correcting biases and handling boundedly rational forecasters.
6. Visualization and communication: interfaces to display and elicit information, and other best practices for communicating uncertainty and educating the public about forecasts.
7. Fielded platforms and systems: forecasting in support of decision making by companies, organizations, or governments.

Call For Participation

We invite both full contributions and poster contributions. A full contribution is an unpublished or recently published research manuscript. A poster contribution can be a preprint, a recently published paper, an abstract, or a presentation file. Preference may be given to more recent and unpublished work. We especially encourage poster contributions from students and postdocs.

Please submit your contributions using the link https://forms.gle/xNbQSrEjxTawRPRB8 by August 2, 2024.  The workshop is non-archival, meaning contributors are free to publish their results later in archival journals or conferences. Email questions or suggestions to the organizers.

The workshop will include invited and contributed talks, open and/or panel discussion, and a poster session. Workshop registration is open to all but you must register to attend.

Important Dates

Submissions due: Friday, August 2, 2024. (AoE)

Notifications: Friday, August 9, 2024

Workshop Date: Monday, October 14, 2024

Parking: All registered participants for the workshop will receive an email with a link to register their car closer to the workshop. If you do not have a Rutgers parking permit and plan to drive to the workshop you must register your car to park.

Presented in association with the SF on Mechanisms & Algorithms to Augment Human Decision Making.